Mechanisms underlying the pathophysiology of minimal change nephrotic syndrome (MCNS), the most frequent of glomerular diseases in children, remain elusive, although recent arguments suggest that T cell dysfunction may be involved in the pathogenesis of this disease. Recently, we reported that activated T cells of these patients display a down-regulation of IL-12R β2 chain, suggesting an early commitment toward Th2 phenotype. In this study, we show that the short form of the proto-oncogene c-maf, a known activator of the IL-4 gene, is highly induced in MCNS T cells during relapse, where it translocates to the nuclear compartment and binds to the DNA responsive element. Unexpectedly, the nuclear localization of c-maf did not promote the IL-4 gene transcription in relapse. Using several approaches, we show in this study that RelA blunts IL-4 induction in T cells during the relapse in these patients. We demonstrate that the ex vivo inhibition of proteasome activity in T cells from relapse, which blocks NF-κB activity, strongly increases the IL-4 mRNA levels. Overexpression of c-maf in T cells induces a high level of IL-4 promoter-driven luciferase activity. In contrast, coexpression of c-maf with NF-κB RelA/p50, or RelA, but not p50, inhibits the c-maf-dependent IL-4 promoter activity. Finally, we demonstrated that, in T cell overexpressing RelA and c-maf, RelA expelled c-maf from its DNA binding site on IL-4 gene promoter, which results in active inhibition of IL-4 gene transcription. Altogether, these results suggest that the involvement of c-maf in Th2 commitment in MCNS operates through IL-4-independent mechanisms.

Minimal change nephrotic syndrome (MCNS) 3 is a glomerular disease characterized by heavy proteinuria with relapse/remission course without renal evidence of classical immune mechanism-mediated injury (1). Although MCNS occurs at all ages, early relapses are most frequent in children and young adults. Relapses are often triggered by upper respiratory tract infections or other immunogenic stimuli such as exposure to an allergen, bee stings, or contact dermatitis (2, 3, 4). In most cases, MCNS is steroid sensitive, but relapses often occur when steroids are withdrawn or decreased. Indirect evidence suggesting a pathogenic lymphocyte factor has come from experiments showing that systemic infusion in rats of supernatants of cultured PBMC or T cells of patients with MCNS relapse induces proteinuria (5, 6, 7). These observations have led to postulation that MCNS results from immune alterations affecting peripheral T cells; however, the molecular link between the immune system and the kidney disease is still lacking.

Under physiological conditions, generation of Th-expressing Th1 or Th2 cytokines following priming of naive CD4+ T cells by Ag, is influenced by the nature of TCR signaling as well as by cytokines locally present during initiation of the T cell response. Th1 cells produce IFN-γ and IL-2, whereas Th2 cells synthesize IL-4, IL-5, and IL-13, among others. Previously, we have demonstrated that T cells from relapse display a down-regulation of the IL-12R β2 chain, which is compatible with Th2 polarization in MCNS (8), in agreement with several reports showing an increase of IL-13 in relapse (9, 10). In addition, patients with MCNS often display a defect in delayed-type hypersensitivity response, suggesting an abnormal Th1-dependent cellular immunity (11).

Among the up-regulated genes in MCNS T cells, we identified c-maf, which is the cellular homolog of the viral oncogene maf, a member of the basic region/leucine zipper transcription factor family (12). The human c-maf gene is located on chromosome 16q22-q23, and transcribed in several mRNA forms, generated by alternative splicing (13). c-maf belongs to the large maf family, which includes NRL (neural retinal-specific gene) and mafB, which exhibit tissue-specific functions. They are characterized by an N-terminal proline/serine/threonine-rich acidic transactivation domain and a C-terminal basic region containing the DNA binding domain. Although maf family members bind to similar DNA motifs, the amino acid sequence adjacent to the DNA binding domain is crucial for the specificity of the target gene recognition. c-maf binds as a homo- or heterodimer to the DNA palindromic Maf recognition element (Mare) that consists of an extended AP1 motif (14). Unlike other transcription factors involved in Th polarization, c-maf displays some specific characteristics. It is barely expressed in resting T cells and is mainly induced by signals emanating from proximal events involving TCR activation (15, 16). In contrast to most transcription factors, the only known direct downstream target of c-maf is the IL-4 gene (15, 17), whereas GATA-3, another major Th2 transcription factor, directs the activation of IL-4, IL-5, and IL-13 genes (18).

In the present study, we show that c-maf is highly induced in MCNS. During relapse, its nuclear localization, together with its binding to the DNA consensus sequence, suggest that c-maf is functionally active. Unexpectedly, very low levels of IL-4 mRNA were detected. In contrast, remissions were associated with a significant increase in IL-4 mRNA level relative to relapse. We have previously shown that peripheral T lymphocytes of patients with MCNS relapse display sustained NF-κB activation, partly mediated by an inappropriate increase in proteasome activity (19). Supporting evidence came from the identification of several genes up-regulated in relapse and closely related to NF-κB activation and the proteasome complex (8). We showed in this study that the lack of IL-4 induction in relapse results from this NF-κB activation involving the NF-κB RelA/p50 or RelA complexes. This functional antagonism between c-maf and RelA might explain why patients with MCNS often display low IL-4 levels during the relapse phase and suggests that c-maf-mediated polarization of MCNS T cells evolves through IL-4-independent mechanisms.

The clinical characteristics of the patients analyzed in this study are summarized in Table I. They include nine children and seven adults with MCNS. Each patient is designed by a fixed number throughout the study. For children, the criteria of the International Study of Kidney Diseases were used for diagnosis and management of MCNS (20). For adults, the diagnosis of MCNS or membranous nephropathy (MN) was confirmed by renal biopsy before inclusion. All patients with relapse (children and adults), as well as patients with MN, had proteinuria >40 mg/kg/24 h, and low serum albumin levels at the time of blood sampling, which was performed before the beginning of steroid treatment. Four children (nos. 1–4) and five adult (nos. 5–9) patients were extensively analyzed in this study, given the availability of enough blood samples for subsequent CD4 T cell subset purification.

Table I.

Clinical parameters of MCNS patientsa

Children with MCNSAdults with MCNSbAdults with MNbNormal ChildrenNormal Adults
Number of patients 
Age (years) 9.7 (3–17) 31 (19–50) 39.6 (28–52) 11.4 (9–17) 24.4 (23–27) 
Boy/girl 4/5 4/3 3/2 3/2 3/1 
Proteinuria (g/day) 8 (3–12) 12 (4–15) 6 (5–15)   
Proteinemia (normal, 65–78 g/L) 43 (39–53) 46 (38–54) 50 (47–56) NT NT 
Serum C-reactive protein (mg/L) 20 (15–30) <15 <15 NT NT 
Plasma creatinine (μM/L) 75 (45–95) 110 (80–130) 120 (95–140) NT NT 
Steroid therapyc Relapse: none Remission: 5 mg alternate day (4 patients) Relapse and remission: none None None None 
Children with MCNSAdults with MCNSbAdults with MNbNormal ChildrenNormal Adults
Number of patients 
Age (years) 9.7 (3–17) 31 (19–50) 39.6 (28–52) 11.4 (9–17) 24.4 (23–27) 
Boy/girl 4/5 4/3 3/2 3/2 3/1 
Proteinuria (g/day) 8 (3–12) 12 (4–15) 6 (5–15)   
Proteinemia (normal, 65–78 g/L) 43 (39–53) 46 (38–54) 50 (47–56) NT NT 
Serum C-reactive protein (mg/L) 20 (15–30) <15 <15 NT NT 
Plasma creatinine (μM/L) 75 (45–95) 110 (80–130) 120 (95–140) NT NT 
Steroid therapyc Relapse: none Remission: 5 mg alternate day (4 patients) Relapse and remission: none None None None 
a

All data are presented as means, and extreme values are in parentheses. NT, Not tested.

b

Diagnosis was carried out by renal biopsy.

c

At the time of blood sampling.

Remission samples were collected during periods of inactive disease, defined by a proteinuria <0.2g/24 h, without steroid treatment (1–3 mo following stopping of steroids), except for patients 1, 3, 7, and 8, who received 5 mg of prednisone on alternate day. Controls consisted of five normal children (nos. 1–5) studied while undergoing routine analysis, four normal adult volunteers (nos. 6–9), as well as five patients with idiopathic MN (MN1–MN5). The idiopathic character of nephrotic syndrome in MN patients was retained when clinical and laboratory investigations excluded any underlying disease such as infectious or malignant process.

An attempt was made to match normal subjects and nephrotic patients by age and sex. Informed consent was obtained from the parents and whenever possible from the pediatric patients, as well as from adult patients and normal volunteers.

Immunoselection of T cell subsets and monocytes was performed using a mixture of hapten-conjugated Abs (Miltenyi Biotec, Auburn, CA). Abs raised against c-maf (M153, sc-7866, recognizing the short and the long forms), NF-κBp50 (sc-7178X), NF-κBp65 (RelA, sc-372X), and SP1 (sc-420) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The M153 polyclonal Ab recognizes the amino acid sequence 19–171, corresponding to the N-terminal transactivation domain of c-maf. A polyclonal Ab specific for c-maf-long (c-maf-L) (see Results) was prepared against a 13-aa peptide (aa 378–391 of the protein) (Eurogentec, Seraing, Belgium). The proteasome inhibitor MG132 (carbobenzoxyl-leucinyl-leucinyl-leucinal-H; Z-LLL) was purchased from Calbiochem (La Jolla, CA).

PBMC were purified from 5-ml (children, <7 years old), 10-ml (children, 8–12 years old), 15-ml (children, 13–17 years old), and 30- to 40-ml (adults) blood samples, in accordance with ethical committee rules. Whenever possible, one-third of the sample was left for PBMC analyses (RNA, protein extracts, and immunohistochemical analyses), whereas two-thirds of PBMC preparations were used for T cell subset isolations as previously described (19). The purity of the preparations was 88–96%, as assessed by flow-cytometric analysis, using FITC-conjugated CD2, CD4, CD19, and CD8 Abs. Cell analyses including RNA and protein experiments were performed without any exogenous stimulation.

PBMC, or T cell subsets, were spread at 105 cells/slide, fixed, and permeabilized by methanol at −20°C, and then processed for immunoreactivity. Cells were incubated in the blocking solution (10% normal sheep serum, 1% BSA) for 40 min, washed twice with PBS, and then incubated with c-maf Ab (10 μg/ml in 5% normal sheep serum, 1% BSA, 0.1% Tween 20) for 2 h at room temperature. Slides were washed three times, and then incubated with anti-rabbit Cy3-labeled Ab (1/1000 in blocking solution) for 30 min. Slides were mounted in a Vectashield 4′,6′-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA), and analyzed on an Axioplan Zeiss (Oberkochen, Germany) microscope equipped for epifluorescence. The percentage of positive cells was determined on an average of 200 cells.

Total RNAs were treated by DNase I and purified using RNeasy kit (Qiagen, Valencia, CA). The absence of genomic DNA contamination was checked by running an aliquot of total RNA samples. The sequence of the primers, and PCR parameters are indicated in Table II. To determine the relative expression of the short (c-maf-short (c-maf-S)) and the long (c-maf-L) forms of c-maf mRNA, we selected two sets of primers. Because the two have the same sequence until the stop codon of the short form (position 1926), we selected an antisense primer in the 3′ untranslated region of c-maf-S and in the second exon of c-maf-L, respectively (Table II). Semiquantitative RT-PCRs were performed as previously described (19). Each cycle consists of a denaturation step at 94°C for 30 s, annealing for 30 s at the indicated temperature (Table II), and extension at 68°C for 2 min. Amplified products were detected on Southern blots with 32P-labeled specific internal oligonucleotide probes and quantified using the PhosphorImager (Storm 840; Molecular Dynamics, Sunnyvale, CA), coupled with the ImageQuant, version 1.11, analysis software. PCR were normalized for GAPDH expression.

Table II.

Sets of primers used in semiquantitative and quantitative RT-PCRa

mRNAOligonucleotidesAccession no.Expected Size (bp)Annealing Temperature (°C)PCR Cycles
GAPDH S: 5′-ACCACAGTCCATGCCATCAC-3′ NM 004048 374 58 25 
 AS: 5′-TCCACCACCCTGTTGCTGTA-3′     
 I: 5′-CTCAAGGGCATCCTGGGCTACACTGAGCAC-3′     
      
c-maf S (position 1571–1592): 5′-TGCACTTCGACGACCGCTTCTC-3′ AF055376/7    
 AS—short form (position 1957–1938): 5′-GGTGGCTAGCTGGAATCGCG-3′ AF055376 386 60 32 
 AS—long form (position 2040–2011): 5′-TGTACAGCTCTCACACAAATTTCATTTTGT-3′ AF055377 469 62 32 
 I: 5′-GCTGCTGCAGCAAGTCGACCACCTC-3     
 CD5: 5′-GAGGCAGGAGGATGGCTTCAGAACTGGCAATGAACAATTCCGACCTGCCCA-3′   72  
 CD3: 5′-CGCGTGTCACACTCACATGAAAAATTCGGGAGAGGAAGGGTTGTCG-3′ AF055376    
      
IL-10 S: 5′-AGTCTGAGAACAGCTGCACCCACTTC-3′ XM001409 215 60 40 
 AS: 5′-GGGCATCACCTCCTCCAGGTAA-3′     
      
IL-4 S: 5′-TTCTCCTGATAAACTAATTGCCTCACATTGTC-3′ XM004053 143 60 40 
 AS: 5′-GGTGATATCGCACTTGTGTCCGTGG-3′     
      
IFN-γ S: 5′-GGTTCTCTTGGCTGTTACTGC-3′ XM006883 294 60 40 
 AS: 5′-GTCATCTCGTTTCTTTTTGTTGCT-3′     
mRNAOligonucleotidesAccession no.Expected Size (bp)Annealing Temperature (°C)PCR Cycles
GAPDH S: 5′-ACCACAGTCCATGCCATCAC-3′ NM 004048 374 58 25 
 AS: 5′-TCCACCACCCTGTTGCTGTA-3′     
 I: 5′-CTCAAGGGCATCCTGGGCTACACTGAGCAC-3′     
      
c-maf S (position 1571–1592): 5′-TGCACTTCGACGACCGCTTCTC-3′ AF055376/7    
 AS—short form (position 1957–1938): 5′-GGTGGCTAGCTGGAATCGCG-3′ AF055376 386 60 32 
 AS—long form (position 2040–2011): 5′-TGTACAGCTCTCACACAAATTTCATTTTGT-3′ AF055377 469 62 32 
 I: 5′-GCTGCTGCAGCAAGTCGACCACCTC-3     
 CD5: 5′-GAGGCAGGAGGATGGCTTCAGAACTGGCAATGAACAATTCCGACCTGCCCA-3′   72  
 CD3: 5′-CGCGTGTCACACTCACATGAAAAATTCGGGAGAGGAAGGGTTGTCG-3′ AF055376    
      
IL-10 S: 5′-AGTCTGAGAACAGCTGCACCCACTTC-3′ XM001409 215 60 40 
 AS: 5′-GGGCATCACCTCCTCCAGGTAA-3′     
      
IL-4 S: 5′-TTCTCCTGATAAACTAATTGCCTCACATTGTC-3′ XM004053 143 60 40 
 AS: 5′-GGTGATATCGCACTTGTGTCCGTGG-3′     
      
IFN-γ S: 5′-GGTTCTCTTGGCTGTTACTGC-3′ XM006883 294 60 40 
 AS: 5′-GTCATCTCGTTTCTTTTTGTTGCT-3′     
a

The oligonucleotides were selected from the sequences with the indicated accession numbers. The size of each amplified product, its annealing temperature, and the number of PCR cycles are indicated. The CD5 and CD3 primers were used to obtain the coding sequence of c-maf short form (see Materials and Methods). S, Sense; AS, antisense; I, internal oligonucleotide probe.

Quantitative RT-PCR was performed using the Light Cycler (Roche Diagnostics, Somerville, NJ), as previously described (19).

Preparation of cytosolic and nuclear extracts and EMSA experiments were performed as already reported (19). The following oligonucleotides and their complementary strands were synthesized (Genset-Proligo, Paris, France): wild-type (wt) Mare, 5′-GGAATTGCTGACTCAGCATTACT-3′; and mutant (mt) Mare, 5′-GGAATTGCTGACTCATTGTTACT-3′; the c-maf consensus recognition sequence is underlined, and the mutated nucleotides are in bold (14). The human IL-4 promoter regions (−95/+55) and (−65/+55) base pair (+1 corresponds to the IL-4 transcription initiation site) were isolated by digestion of IL-4 promoter-containing plasmids (a gift of Dr. V. Casolaro (The Johns Hopkins University, Baltimore, MD)) with HindIII-XbaI restriction enzymes. The wild and mutant SP1 oligonucleotides were obtained from Santa Cruz Biotechnology. The IL-4 probes (25 ng) were labeled by random hexamer priming using [α-32P]dCTP and the klenow fragment of DNA polymerase I (Life Technologies, Eragny, France), whereas the Mare and SP1 oligonucleotides were 5′-end labeled using [γ-32P]ATP (3000 Ci/mmol; Amersham Pharmacia Biotech, Paisley, U.K.) and T4 polynucleotide kinase (New England Biolabs, Hertfordshire, U.K.). Radiolabeled products were purified on STE-10 or STE-30 spin column chromatography (Clontech, Palo Alto, CA). Specificity of the Mare and SP1 binding was tested by using mutant probes, whereas the binding to IL-4 promoter was analyzed by cold competition using an oligonucleotide spanning bp −87 to −38 that includes the NF-κB and Mare (underlined) putative sites (wt IL-4 −87/−38, 5′-GTGT AACGA AAA T T T CCAA T G T AAAC T CA T T T T CCC T CGG TTTCAGCAAT-3′), respectively. In some experiments, the potential binding of nuclear extracts from MCNS T cells to IL-4 promoter was tested by using the wt IL-4 −87/−38 and its mutated form in the NF-κB site (mt IL-4 −87/−38, 5′-GTGTAACGGCAACCTCCAATGTAAACTCATTTTCCCTCGGTTTCAGCAAT-3′). The presence of c-maf in DNA-protein complexes was determined by preincubation of nuclear extracts with 2 μg of specific polyclonal Ab (M153) before the addition of the probe. Gels were dried and revealed after overnight exposition on a PhosphorImager (Storm 840; Molecular Dynamics). Western blotting was performed as previously described (19).

CD4 T cell fractions of MCNS relapse were subdivided into two identical fractions and incubated overnight, in the absence (control) or in the presence of 10 μM MG132, and then processed for RNA and protein extraction as previously described (19).

The IL-4 promoter fragments corresponding to −95 and −65 bp were dephosphorylated with calf intestinal alkaline phosphatase (Promega, Paris, France), and then inserted in the Bluescript plasmid (Stratagene, La Jolla, CA) at the indicated sites. The IL-4 promoter inserts were further isolated by digestion with KpnI-SacI and subcloned in pGL2 enhancer reporter plasmids (Promega). NF-κBp65 and -p50 expression vectors have been described (21). The c-maf-S coding sequence (1119 bp) was isolated by digestion with BamHI-HindIII restriction enzymes of a 4-kb c-maf cDNA-containing plasmid (gift of Q. Chen (Baylor College of Medicine, Houston, TX) (22)). The BamHI-HindIII fragment was sequentially digested with ApoI and BsrDI, and then purified on STE 100 chromaspin column. The ApoI-BsrDI fragment was amplified by PCR, using sense (CD5) and antisense (CD3) primers (indicated in Table I), under the following conditions: denaturation at 95°C for 1 min, 35 two-step cycles (95°C for 15 s, 72°C for 3 min), and final extension at 72°C for 3 min. The PCR product was ligated into pcDNA3.1/V5-His-Topo expression vector (Invitrogen, Paisley, U.K.). To minimize variations in transfection efficiency, the same cell passage numbers were used for all transfections. Cells were cultured in RPMI 1640 medium supplemented with 10% FBS and antibiotics (penicillin, streptomycin), and maintained at 37°C in an atmosphere of 5% carbon dioxide. Jurkat T cells (107 per condition) were transiently transfected with IL-4-Luc reporter plasmid (0.5 μg/106 cells) and expression plasmid (2 μg/106 cells) by electroporation using a Bio-Rad (Paris, France) gene pulser set at 960 μF and 250 V. Cells were allowed to recover overnight and harvested 18-h posttransfection. Luciferase activity was determined in 10 μl of cell extracts using the luciferase assay substrate (Promega) with a Lumat LB 9507 luminometer (PerkinElmer, Weiterstadt, Germany) and normalized for protein content measured by the Bradford reagent (Bio-Rad). Transfections were performed in duplicate, and the results of at least four independent experiments were calculated as mean ± SD values for luciferase activity.

Results were expressed as mean ± SD. Statistical significance was determined using the Mann-Whitney nonparametric test when three or more groups were analyzed or, if less, the parametric Fisher test, according to data form. Differences were considered significant at a level of p < 0.05

Pursuant to our earlier observations suggesting early commitment to the Th2 lineage in patients with MCNS relapses and an increase in c-maf transcript during relapse vs remission, we assessed c-maf protein in nuclear extracts from CD4+ T cells derived from patients during relapse and remission phases. In the four patients tested, c-maf was detected in nuclear extract from relapse, but no signal was apparent in remission (Fig. 1). The presence of c-maf in nuclei does not prove that the protein is functional. To address this question, we analyzed, by EMSA experiments, c-maf-dependent DNA binding activity in CD4 T cells in seven patients with MCNS relapse, and in six of them in remission (Fig. 2,A). The higher band shifts, detected in nuclear extracts of CD4+ T cells from relapse, corresponded to Mare binding activity, as attested by the loss of these complexes in the presence of mutant Mare oligonucleotide. The significance of the lower migrating complexes was undetermined. No band shift was detected in remission. To control that the lack of c-maf DNA-binding activity in remission did not result from nuclear protein degradation, the same extracts were analyzed with a SP1 probe. SP1 binding motifs are frequently recruited to initiate transcription of TATA-box-less genes, including many housekeeping genes, so that nuclear extracts typically exhibit several band shifts corresponding to SP1-dependent DNA binding domains (23). The presence of multiple SP1-band shifts in remission samples confirmed their integrity and clearly established that the lack of Mare DNA binding activity in remission resulted from the exclusion of c-maf from nuclear compartment. Preincubation of nuclear extracts of CD4+ T cells from relapse with anti-c-maf Ab (M153) significantly reduced the band shifts (Fig. 2 B), suggesting that the Ab interferes with the binding of c-maf to its responsive site and does not induce a supershift band, as reported by others (24).

FIGURE 1.

Expression of c-maf protein in CD4+ T cells in MCNS relapse. Nuclear protein extracts (20 μg) of CD4+ T cells from relapse and remission were analyzed by Western blot using c-maf Ab (M153). The blots were subsequently stripped and reprobed with anti-SP1 Ab.

FIGURE 1.

Expression of c-maf protein in CD4+ T cells in MCNS relapse. Nuclear protein extracts (20 μg) of CD4+ T cells from relapse and remission were analyzed by Western blot using c-maf Ab (M153). The blots were subsequently stripped and reprobed with anti-SP1 Ab.

Close modal
FIGURE 2.

c-maf DNA binding activity in MCNS relapse. A, EMSA analyses for c-maf-dependent DNA binding in CD4+ T cell from seven patients in nephrotic relapse (Rel) and six of them in remission; nuclear extracts (10 μg) were tested for shift assays using the wt Mare. The specificity of the band shift was demonstrated by the loss of this band in the presence of the mt Mare. The integrity of the nuclear extracts was tested by shift assays using SP1 oligonucleotide. The asterisk indicates lower migrating complexes of undetermined significance. B, Alteration of c-maf binding to Mare site in the presence of anti-c-maf Ab (M153). EMSA analyses of nuclear extracts from MCNS relapse, for Mare site, in the absence (−) or the presence (+) of Ab.

FIGURE 2.

c-maf DNA binding activity in MCNS relapse. A, EMSA analyses for c-maf-dependent DNA binding in CD4+ T cell from seven patients in nephrotic relapse (Rel) and six of them in remission; nuclear extracts (10 μg) were tested for shift assays using the wt Mare. The specificity of the band shift was demonstrated by the loss of this band in the presence of the mt Mare. The integrity of the nuclear extracts was tested by shift assays using SP1 oligonucleotide. The asterisk indicates lower migrating complexes of undetermined significance. B, Alteration of c-maf binding to Mare site in the presence of anti-c-maf Ab (M153). EMSA analyses of nuclear extracts from MCNS relapse, for Mare site, in the absence (−) or the presence (+) of Ab.

Close modal

To analyze the expression profile of c-maf protein during the disease, we immunoblotted cytoplasmic extracts from PBMC of nine patients in relapse and in remission phase. We detected a strongly expressed band with an apparent molecular mass of 50 kDa and a weaker band at 55 kDa (Fig. 3,A). The 50-kDa band was barely detected in cytoplasm extracts from relapse in five patients (nos. 1–4 and 6) but it was more evident in four patients (nos. 5 and 7–9). In contrast, this protein band was prominent in the cytoplasmic extracts of the nine patients in remission. Because nuclear extracts of these same cells contain a high amount of c-maf (Fig. 1), we deduced that c-maf translocates from nuclear to cytoplasmic compartment during the remission phase. The 55-kDa band exhibited a similar pattern, except for patients 4 and 6 in whom it was visualized in the absence of the major form in the cytoplasm extract from relapse.

FIGURE 3.

Immunological detection of c-maf. A, expression of c-maf protein in MCNS during the relapse and the remission phases. Cytoplasmic protein extracts (50 μg) of PBMC were analyzed by Western blotting using anti-c-maf Ab (M153). B, Comparative expression between MCNS patients and controls. Equal amount of protein extracts (50–60 μg) from PBMC of seven patients with MCNS in remission (cytosolic fractions), three nephrotic patients with MN, and nine normal subjects (N) (whole-cell lysates) were analyzed by Western blotting using the M153 Ab. The position of c-maf (including the doublet when discernible) is indicated. The blots A and B were subsequently stripped and reprobed with anti-actin Ab.

FIGURE 3.

Immunological detection of c-maf. A, expression of c-maf protein in MCNS during the relapse and the remission phases. Cytoplasmic protein extracts (50 μg) of PBMC were analyzed by Western blotting using anti-c-maf Ab (M153). B, Comparative expression between MCNS patients and controls. Equal amount of protein extracts (50–60 μg) from PBMC of seven patients with MCNS in remission (cytosolic fractions), three nephrotic patients with MN, and nine normal subjects (N) (whole-cell lysates) were analyzed by Western blotting using the M153 Ab. The position of c-maf (including the doublet when discernible) is indicated. The blots A and B were subsequently stripped and reprobed with anti-actin Ab.

Close modal

Given the high expression of c-maf in MCNS relapse, we wanted to determine its basal expression in nine normal subjects (five children, N1–N5; four adults, N6–N9). In parallel, to ascertain whether c-maf was not up-regulated in response to protein leakage, three patients with nephrotic syndrome linked to MN were included for analysis. In five of nine normal subjects, c-maf protein was below the detection limits, whereas in the remaining four patients, the 55-kDa band was detected at lower level except for subject 7 who displayed a significant expression (Fig. 3 B). It is noteworthy that the 50-kDa band was not detected in normal subjects as well as in nephrotic patients with MN for whom only the 55 kDa-band was visualized.

We analyzed the cellular distribution of c-maf and the frequency of positive cells in relapse and in remission. The expression of c-maf in CD4+ T cells is showed for two patients, no. 3 (43-year old) and no. 8 (14-year old) (Fig. 4). During relapse, c-maf was visualized essentially inside the nuclear compartment of CD4+ T cells, and the nuclear localization in positive cells was confirmed by double staining with the DAPI dye. In remission, c-maf was detected in the peripheral area, outside of the DAPI staining, compatible with the cytoplasmic localization. Data obtained in 16 patients with MCNS, 5 MN patients and 6 normal controls are summarized in Table III. Two hundred cells per sample were analyzed. Nuclear c-maf expression was seen in 5–7% PBMC and 15–20% of CD4+ T cells during the relapse, whereas in remission c-maf was detected in cytoplasm of 10–15% PBMC and 25% CD4+. These results were observed when patients were examined during their first flare, but the percentage of positive cells seems more important in frequent relapsers (data not shown). In contrast, c-maf was consistently absent from nuclear area and very modestly or not detected in cytoplasm of PBMC, in MN patients, or normal subjects.

FIGURE 4.

Immunofluorescence detection of c-maf. Upper panel, Immunostaining of CD4 T cells of patient 3 (43 years old) (first relapse; magnification, ×40). Lower panel, Immunostaining of CD4 T cells of patient 8 (14 years old) (first relapse; magnification, ×40).

FIGURE 4.

Immunofluorescence detection of c-maf. Upper panel, Immunostaining of CD4 T cells of patient 3 (43 years old) (first relapse; magnification, ×40). Lower panel, Immunostaining of CD4 T cells of patient 8 (14 years old) (first relapse; magnification, ×40).

Close modal
Table III.

Cellular distribution and percentage of c-mafpositive cells in 16 patients with MCNS (first flare), 5 patients with MN, and 6 normal subjectsa

RelapseRemissionMNNormal Subjects
Nuclear stainingNuclear stainingCytoplasm staining
PBMC 5–7% ND 10–15% ND—trace ND—trace 
CD4+ T cells 15–20% ND 25% ND ND 
RelapseRemissionMNNormal Subjects
Nuclear stainingNuclear stainingCytoplasm staining
PBMC 5–7% ND 10–15% ND—trace ND—trace 
CD4+ T cells 15–20% ND 25% ND ND 
a

Two hundred cells were analyzed for each measurement.

Recently, two c-maf mRNAs have been identified (13). The short form (c-maf-S) corresponds to an intronless genomic sequence of 4248 bp encoding a predicted protein of 373 aa. The long form (c-maf-L) is generated by alternative splicing (Fig. 5,A). A short exon (220 bp) located 1.9 kb downstream from the c-maf-S sequence, is inserted at the position 1925 in the open reading frame of the c-maf-S. This splicing lengthens the coding sequence c-maf-L of 30 aa. To assess whether the 50-kDa protein band detected by Western blotting corresponds to c-maf-S, we investigated by RT-PCR the relative expression level of both transcripts in PBMC from relapse and remission phases in nine patients, using specific primers. We found a major induction of c-maf-S mRNA during relapse as compared with remission (Fig. 5 B). In contrast, the expression of c-maf-L transcript was significantly expressed in patients 5 and 9 during the relapse and in patient 8 during the remission, whereas it was barely detectable in others. We concluded that the 50-kDa protein band identified in Western blotting likely corresponds to c-maf-S. Interestingly, we note that the expression of c-maf-S declines in most patients during the remission phase. This result suggests that the c-maf protein visualized in cytoplasmic compartment during the remission mainly arises from c-maf produced during the relapse phase. In contrast, normal subjects exhibited very modest basal levels of c-maf mRNA, in accordance with Western blotting data. The strong induction of c-maf mRNA appeared specific to MCNS, because MN patients suffering a similar range of proteinuria exhibited a c-maf mRNA level undistinguishable from that of normal subjects.

FIGURE 5.

Differential expression of c-maf-S and c-maf-L mRNAs in MCNS. A, Schematic organization of the c-maf-S and c-maf-L. A single exon encodes the c-maf-S, whereas two exons 1919 bp apart encode the c-maf-L. B, RT-PCR analyses using a sense primer common to c-maf-S and c-maf-L, and distinct antisense primers (Table II). After Southern blotting, the PCR products were detected with specific internal oligonucleotide probes (I). The expression of GAPDH was monitored in parallel, to control the initial amount of mRNA. The lower panel shows the quantification of PCR products as determined using the Image Quant, version 1.11, analysis software, after normalization against the corresponding GAPDH mRNA values (▪, relapse; □, remission). C, Maximal induction of c-maf in the CD4+ T cell subset. RT-PCR analyses for c-maf expression in nine patients. CD4+ and non-CD4+ T cells fractions were purified from PBMC as described in Materials and Methods. The expression of GAPDH was monitored in parallel. Shown is densitometric analysis of data, using the Image Quant, version 1.11, analysis software, after normalization against the corresponding GAPDH mRNA. Statistics represent the mean ± SD.

FIGURE 5.

Differential expression of c-maf-S and c-maf-L mRNAs in MCNS. A, Schematic organization of the c-maf-S and c-maf-L. A single exon encodes the c-maf-S, whereas two exons 1919 bp apart encode the c-maf-L. B, RT-PCR analyses using a sense primer common to c-maf-S and c-maf-L, and distinct antisense primers (Table II). After Southern blotting, the PCR products were detected with specific internal oligonucleotide probes (I). The expression of GAPDH was monitored in parallel, to control the initial amount of mRNA. The lower panel shows the quantification of PCR products as determined using the Image Quant, version 1.11, analysis software, after normalization against the corresponding GAPDH mRNA values (▪, relapse; □, remission). C, Maximal induction of c-maf in the CD4+ T cell subset. RT-PCR analyses for c-maf expression in nine patients. CD4+ and non-CD4+ T cells fractions were purified from PBMC as described in Materials and Methods. The expression of GAPDH was monitored in parallel. Shown is densitometric analysis of data, using the Image Quant, version 1.11, analysis software, after normalization against the corresponding GAPDH mRNA. Statistics represent the mean ± SD.

Close modal

To determine the cellular source of the c-maf-S transcript, we analyzed its expression level in CD4+ T cells purified by negative immunomagnetic selection from PBMC of six adult and three children patients with relapse. The highest level of c-maf mRNA was observed in the CD4+ T cell subset as compared with the non-CD4+ T cell fractions (Fig. 5 C). We concluded that the burst of c-maf induction was predominant in CD4+ T cells.

In T cells, the IL-4 gene, but not other Th2 cytokine genes, is a well-known downstream target of the c-maf transcription factor (15, 17). Given the nuclear localization and the DNA binding activity of c-maf, we speculated that MCNS relapses were associated with up-regulation of IL-4 and inhibition of Th1 cytokines. Thus, we determined in the same cohort of patients the level of IL-4, IFN-γ, and IL-10 cytokine transcripts by RT-PCR using total RNA from PBMC without exogenous cell stimulation. Surprisingly, we found that the expression of IL-4 transcript was very low in four patients, in both relapse and remission phases, and furthermore, in five (nos. 2, 5, 7, 8, 9), a down-regulation of IL-4 mRNA was observed during relapse, relative to remission (Fig. 6,A). The difference between IL-4 mRNA expression in relapse and in remission was significant (p < 0.05), but there was no significant difference in the IL-4 level between the remission phases and normal subjects. In addition, there was no significant difference in IFN-γ, and IL-10 mRNA expression in relapse and remission phases as compared with MN patients and normal controls (Fig. 6 B).

FIGURE 6.

A, Decoupling between c-maf induction and IL-4 mRNA expression in MCNS. The upper panel represents quantitative RT-PCR analyses, from PBMC total RNA, for IL-4 mRNA during the relapse (▪) and the remission (□) phases. Results are expressed as copy number of IL-4 for 106 copies of GAPDH transcripts. The lower panel represents the quantification of c-maf-S PCR such as depicted in Fig. 4 B. B, Relative expression of IL-4 (from data in A), IL-10, and IFN-γ in patients with MCNS (n = 9) relapse (Rel) and remission (Rem), and MN (n = 5), and in normal controls (CON; n = 9). ∗, p < 0.05 (Mann-Whitney test).

FIGURE 6.

A, Decoupling between c-maf induction and IL-4 mRNA expression in MCNS. The upper panel represents quantitative RT-PCR analyses, from PBMC total RNA, for IL-4 mRNA during the relapse (▪) and the remission (□) phases. Results are expressed as copy number of IL-4 for 106 copies of GAPDH transcripts. The lower panel represents the quantification of c-maf-S PCR such as depicted in Fig. 4 B. B, Relative expression of IL-4 (from data in A), IL-10, and IFN-γ in patients with MCNS (n = 9) relapse (Rel) and remission (Rem), and MN (n = 5), and in normal controls (CON; n = 9). ∗, p < 0.05 (Mann-Whitney test).

Close modal

Given the strong induction of c-maf during the relapse, the down-regulation of IL-4 was unexpected, but it was in agreement with most of the reports showing a lack of increase of IL-4 in MCNS patients (10, 25, 26, 27). We have previously reported that MCNS relapses are associated with a sustained NF-κB activation with an increase of proteasome activity, compatible with a breakdown in the autoregulatory feedback loop (19). We hypothesized that the high NF-κB activity might counteract the transactivation of the IL-4 gene by c-maf. This suggestion was based on the fact that the IL-4 promoter contains, close to the c-maf response element (bp −45 to −40), a NF-κB binding site (bp −80 to −71). To address this question, we used several approaches. First, we blocked the NF-κB activation in CD4+ T cells from relapse by inhibiting the proteasome activity using the tripeptide aldehyde reagent MG132. We found that the inhibition of the proteasomal activity induced a strong increase of IL-4 transcript as compared with MG132-untreated relapse (Fig. 7,A). This result may be explained by the restoration of IκBα by MG132, thus preventing NF-κB activation, as we have previously reported (19). Second, transient transfection experiments were conducted in Jurkat T cells using the IL-4-luciferase reporter constructs containing c-maf alone (IL-4 −65) or c-maf and NF-κB responsive elements (IL-4 −95). Jurkat T cells were cotransfected with the expression vectors containing the coding sequence of c-maf alone, or with the expression vectors coding for the subunits NF-κBp50 and NF-κBp65 (RelA). Overexpression of c-maf induced an activation of the IL-4 −95 promoter construct (Fig. 7,B). In contrast, cotransfection with RelA alone or RelA/p50 heterodimer expression vectors resulted in down-regulation of the IL-4 promoter activity. Surprisingly, cotransfection with NF-κBp50 significantly enhances the c-maf-induced IL-4 promoter activity. In contrast, deletion of the binding site for NF-κB in the IL-4 promoter construct (IL-4 −65), reduced the c-maf-induced luciferase activity by 50% and removes the influence of NF-κB proteins on the reporter activity (Fig. 7 C). These results suggest that increased RelA, whether as homodimers or as heterodimers with p50, significantly blunts the induction of IL-4 by c-maf, whereas the NF-κBp50 homodimers have rather an agonistic effect. We conclude that the NF-κB binding site is required for modulation of c-maf activity on IL-4 gene.

FIGURE 7.

Effect of NF-κB on the IL-4 induction by c-maf. A, Proteasome inhibition induces an increase of IL-4 mRNA expression. CD4 T cells from five patients with relapse and four normal subjects were subdivided into two equal fractions and incubated overnight in the presence (+) or in the absence (−) of the proteasome inhibitor MG132. Total RNA was extracted and used for RT-quantitative PCR analyses for IL-4 mRNA expression. B and C, Cotransfection of T cell Jurkat with NF-κB expression vectors results in differential modulation of c-maf-mediated IL-4 promoter activity. An IL-4 promoter construct containing (IL-4 −95) (B) or lacking (IL-4 −65) (C) the NF-κB binding site was inserted in frame in PGL2 luciferase reporter and transiently transfected with expression plasmids as indicated. Luciferase activity was measured 18 h posttransfection on cell extracts and normalized by protein content determined by using a Bradford assay. These data summarize four experiments for IL-4 −95 and IL-4 −65 promoter constructs.

FIGURE 7.

Effect of NF-κB on the IL-4 induction by c-maf. A, Proteasome inhibition induces an increase of IL-4 mRNA expression. CD4 T cells from five patients with relapse and four normal subjects were subdivided into two equal fractions and incubated overnight in the presence (+) or in the absence (−) of the proteasome inhibitor MG132. Total RNA was extracted and used for RT-quantitative PCR analyses for IL-4 mRNA expression. B and C, Cotransfection of T cell Jurkat with NF-κB expression vectors results in differential modulation of c-maf-mediated IL-4 promoter activity. An IL-4 promoter construct containing (IL-4 −95) (B) or lacking (IL-4 −65) (C) the NF-κB binding site was inserted in frame in PGL2 luciferase reporter and transiently transfected with expression plasmids as indicated. Luciferase activity was measured 18 h posttransfection on cell extracts and normalized by protein content determined by using a Bradford assay. These data summarize four experiments for IL-4 −95 and IL-4 −65 promoter constructs.

Close modal

We further analyzed whether c-maf and NF-κB overexpressions did affect the transcription of the endogenous IL-4 gene, which is embedded in chromatin structure, contrary to reporter gene. To this end, we cotransfected Jurkat T cells with the expression vectors, alone or in combination. Results are depicted in Fig. 8. Overexpression of c-maf alone resulted in active transcription of the endogenous IL-4 gene as attested by the increased IL-4 mRNA level (lane 3), which was strongly inhibited in T cells cotransfected with equal amounts of c-maf and RelA (lane 4). By contrast, the IL-4 transcription gene was amplified by concomitant cotransfection of NF-κBp50 and c-maf (lane 5). Interestingly, forced expression of RelA alone did not influence the IL-4 gene expression, but the transcription of IFN-γ gene was strongly induced (lane 6). It is noteworthy that the effect of RelA on IFN-γ transcription was blunted in T cells overexpressing both RelA and c-maf. In contrast, T cells transfected with NF-κBp50 alone exhibited a high rate of IL-4 gene transcription, whereas the expression of IFN-γ was slightly increased (lane 7). These results support the finding that RelA and c-maf are functionally antagonists, whereas NF-kBp50 and c-maf positively regulate the IL-4 gene expression.

FIGURE 8.

Effects of overexpression of c-maf, NF-κBp50, and RelA on transcription of IL-4 and IFN-γ endogenous genes. T cell Jurkat were transiently transfected with expression plasmids as indicated. Total RNA was extracted 18 h posttransfection and used for RT-quantitative PCR analyses for IL-4 and IFN-γ mRNA expressions.

FIGURE 8.

Effects of overexpression of c-maf, NF-κBp50, and RelA on transcription of IL-4 and IFN-γ endogenous genes. T cell Jurkat were transiently transfected with expression plasmids as indicated. Total RNA was extracted 18 h posttransfection and used for RT-quantitative PCR analyses for IL-4 and IFN-γ mRNA expressions.

Close modal

To understand the mechanisms by which NF-κB exerts a dual effect on the IL-4 induction by c-maf, we analyzed by EMSA the effects of NF-κBp50 or RelA overexpression on c-maf-induced DNA binding complexes. We used as probes promoter regions containing putative c-maf alone (IL-4 −65) or NF-κB and c-maf sites (IL-4 −95). By using the IL-4 −65 probe and nuclear extracts from c-maf-overexpressing T cells, we mainly detected three DNA band shifts, two of which (CI, CIII) were significantly reduced upon incubation with anti-c-maf Ab and virtually disappeared in the presence of a 50-fold excess of the cold Mare oligonucleotide (Fig. 9,A). The complexes CI and CIII were prominent with the IL-4 −95 probe, and were partially neutralized by the anti-c-maf Ab (Fig. 9,B). Interestingly, preincubation with the cold Mare oligonucleotide in excess abrogated the complex CI, suggesting that the complexes CI contained likely only c-maf proteins. Nuclear extracts from NF-κBp50 and RelA overexpressing T cells exhibited also multiple DNA band shifts with the IL-4 −65 probe, but no complex was uppershifted nor showed a lower binding upon incubation with anti-NF-κBp50 or RelA Ab, supporting the fact that the NF-κB-responsive element was located upstream of this promoter region (data not shown). Incubation of nuclear extracts from NF-κBp50-overexpressing T cells with the IL-4 −95 probe induced a higher migrating complex, which was clearly uppershifted with the anti-NF-κBp50 Ab, as well as a lower mobility complex (Fig. 9,C). Nuclear extracts from T cells cotransfected with the RelA expression plasmid produced two distinct band shifts that were strongly reduced in the presence of anti-RelA, suggesting that the binding of RelA to its responsive site was altered by the Ab (Fig. 9,D). Nuclear extracts from T cells transfected with c-maf and NF-κBp50 or c-maf and RelA expression plasmids exhibited a distinct pattern. Forced expression of NF-κBp50 concomitantly with c-maf increased the formation of c-maf-DNA complexes, as compared with overexpression of c-maf alone (Fig. 9, E and B, respectively). The binding of the complexes CI and CIII to IL-4 −95 probe was reduced in the presence of anti-c-maf Ab. Preincubation with the NF-κBp50 Ab did not induce an uppershift, as compared with transient transfection of NF-κBp50 alone (Fig. 9,C), but partially neutralized the DNA binding, suggesting that conformation of the p50 protein in the DNA-protein complexes was modified in presence of the c-maf protein. Moreover, the anti-p50 Ab abrogated the formation of the CIII complex and considerably reduced the intensity of CI complex, suggesting that NF-κBp50 enhances the c-maf DNA binding to IL-4 promoter. The specificity of the binding was demonstrated by the loss of shift complexes in the presence of a 25× excess of cold probe containing the NF-κB and Mare sites (Fig. 9 E, lane 4). Importantly, the binding of c-maf or RelA to their respective responsive sites was abolished in T cells cotransfected with RelA and c-maf. These results suggest that RelA and c-maf are mutually exclusive for their binding to IL-4 promoter.

FIGURE 9.

A, Effects of NF-κB overexpression on c-maf binding to IL-4 promoter site. Nuclear extracts from T cell Jurkat transiently transfected with expression plasmids, as indicated, were tested by EMSA analyzes for binding to IL-4 −65 (lane A) and to IL-4 −95 probes (lanes B–E), respectively. Specificity of band shifts was tested by competition with cold probes, and the nature of IL-4-binding proteins was analyzed by preincubating nuclear extracts with Ab, as indicated. The arrows point to specific complexes (CI–CIII) generated during the assays. Shown is one representative experiment of the five performed with different T cell nuclear extracts. B, DNA binding assays of nuclear extracts from MCNS relapse and remission on the IL-4 −87/−38 probe. T cell nuclear extracts (15 μg) prepared from four patients (nos. 1–4) during the relapse and the remission phase were incubated with the IL-4 probe containing the putative NF-κB and Mare sites (IL-4 −87/−38) in the absence (lanes 1 and 6) or in the presence of anti-RelA (lanes 2 and 7), anti-NF-kBp50 (lanes 3 and 8), and anti-c-maf (lanes 4 and 9) Ab. In lane 5, relapse extracts of patient 4 were incubated with the IL-4 −87/−38 mutated in its putative NF-κB site (mt IL-4 −87/−38).

FIGURE 9.

A, Effects of NF-κB overexpression on c-maf binding to IL-4 promoter site. Nuclear extracts from T cell Jurkat transiently transfected with expression plasmids, as indicated, were tested by EMSA analyzes for binding to IL-4 −65 (lane A) and to IL-4 −95 probes (lanes B–E), respectively. Specificity of band shifts was tested by competition with cold probes, and the nature of IL-4-binding proteins was analyzed by preincubating nuclear extracts with Ab, as indicated. The arrows point to specific complexes (CI–CIII) generated during the assays. Shown is one representative experiment of the five performed with different T cell nuclear extracts. B, DNA binding assays of nuclear extracts from MCNS relapse and remission on the IL-4 −87/−38 probe. T cell nuclear extracts (15 μg) prepared from four patients (nos. 1–4) during the relapse and the remission phase were incubated with the IL-4 probe containing the putative NF-κB and Mare sites (IL-4 −87/−38) in the absence (lanes 1 and 6) or in the presence of anti-RelA (lanes 2 and 7), anti-NF-kBp50 (lanes 3 and 8), and anti-c-maf (lanes 4 and 9) Ab. In lane 5, relapse extracts of patient 4 were incubated with the IL-4 −87/−38 mutated in its putative NF-κB site (mt IL-4 −87/−38).

Close modal

We have previously reported that MCNS relapses are associated with potent NF-κB activation involving the complexes RelA/p50 heterodimers and/or RelA homodimers. Results of transient transfections indicated that RelA might prevent binding of c-maf to its responsive element on IL-4 promoter. We look to determine the pathophysiological relevance of these data by analyzing the potential binding of nuclear extract from MCNS patients (nos. 1–4) during the relapse and the remission phase, on IL-4 promoter. We performed EMSA experiments using the IL-4 −87/−38 probe containing the putative NF-κB and Mare sites. A large band shift with lower mobility was detected in relapse extracts, but no alteration of the binding was observed with anti-RelA, anti-NF-κBp50, nor anti-c-maf Ab (Fig. 9,B, lanes 1–4). Moreover, the mutation of the NF-κB site has little effect on this band shift (Fig. 9 B, lane 5). These results suggest that the band shift observed with relapse extracts and the IL-4 −87/−38 probe did not correspond nor to RelA/p50, nor c-maf proteins, and it likely reflects the binding of other transcription factor(s). We concluded that the presence of RelA (or RelA/p50) in nuclear extracts from relapse might account for the lack of c-maf binding to its IL-4 promoter responsive site. In contrast, only a higher and slight mobility complex was observed in remission extract, which was not altered by anti-c-maf and of which the significance was not determined.

In the present study, we demonstrated for the first time that the short form of c-maf transcript is primarily induced in CD4+ T cells of patients with MCNS relapse, and its expression progressively falls down in remission. c-maf is very modestly expressed in PBMC of normal subjects as well as in patients with MN, suggesting that its induction in MCNS is not a consequence of nephrotic syndrome.

During the relapse phase, c-maf displays functional characteristics evidenced by its nuclear localization and its DNA binding activity. In contrast, during the remission, c-maf protein was only visualized in cytosolic compartment. This finding suggests that the major cytosolic part of c-maf results from its nuclear translocation rather that its continuous active transcription. The persistence of c-maf protein in remission may reflect an increase of protein stability resulting from inhibition of proteasome activity mediated by steroid therapy. A direct effect of glucocorticoids on the down-regulation of c-maf mRNA in remission is unlikely because patients with relapse on steroids exhibit high levels of this transcript (data not shown). In contrast, we cannot exclude that the exclusion of c-maf from nuclear compartment during the remission phase could result from steroid therapy. Altogether, these data raise the possibility that the induction of c-maf transcript is correlated with the activity of MCNS.

The specific induction of c-maf-S as compared with c-maf-L is an intriguing finding with respect to initiating mechanisms of MCNS. In fact, c-maf-S is barely detectable in normal mononuclear cells, and its induction results from proximal signaling events occurring during TCR activation (16, 28). c-maf-S is apparently not regulated by cytokines, whereas c-maf-L is weakly expressed in normal subjects and induced by cytokines (29).

Recently, we found that T cells from MCNS relapse do not display DNA binding activities compatible with the transactivation of IFN-γ (19) and do not express the β2 chain of IL-12R (8), suggesting that MCNS T cells are polarized toward the Th2 pathway. This is relevant to the fact that c-maf is selectively expressed in Th2 cells and inhibits the polarization of naive Th cells along the Th1 pathway through IL-4-dependent and -independent mechanisms (30).

Contrasting with the induction of c-maf, we showed that MCNS relapses are associated with low levels of IL-4, a target gene of c-maf, whereas no significant changes in IL-10 and IFN γ expression between relapse and remission were observed. Although unexpected, the low expression of IL-4 at the mRNA and protein levels in MCNS relapse has been already reported by other investigators (10, 25, 26, 27). In a previous study, we revealed that CD4+ T cells display sustained NF-κB activation during relapse, involving the complexes RelA/p50 heterodimers or (RelA)2 homodimers (19). In the present work, we demonstrate that this sustained NF-κB activation likely antagonizes the transactivation of the IL-4 gene by c-maf. First, we showed that the inhibition of NF-κB activation in T cells from relapse induces a high level of IL-4 such as was expected from c-maf levels in MCNS. Second, we demonstrated that overexpression of RelA alone or Rel A/p50 inhibits the IL-4 promoter activity induced by c-maf. By contrast, overexpression of NF-κBp50 alone enhances the IL-4 promoter activity induced by c-maf. The importance of NF-κBp50 is highlighted by the fact that deletion of the NF-κB binding site results in significant reduction of IL-4 promoter activity. The yield of luciferase activities is relatively low, possibly due to poor stability of the IL-4 reporter plasmid. However, the effects of transfection of expression plasmids on IL-4 endogenous gene transcription are demonstrative. These results suggest that the NF-κB binding site plays a critical and two opposite roles on the regulation of the IL-4 gene. Although the IL-4 levels in remission are not significantly different from those of normal subjects, they are increased during the remission relatively to relapse. The recovery of a normal IL-4 expression level in remission is likely a consequence of steroid therapy, which represses the NF-κB activation by several pathways (31). Indeed, the influence of NF-κB activation on Th phenotype seems to depend on the nature of the NF-κB complexes. Our results clearly suggest that NF-κBp50 enhances the ability of c-maf to bind the IL-4 promoter. In contrast, RelA prevents c-maf binding to its responsive site and conversely. The mechanism of this apparent mutual expelling, not reported so far, remains unclear. Given the close proximity of both NF-κB and Mare sites, a cross-inhibition of the binding by steric hindrance might be postulated. This hypothesis does not explain why NF-κBp50 increases the c-maf binding on IL-4 promoter. A possible explanation is that RelA prevents the binding of endogenous NF-κBp50 to IL-4 promoter and thus counteracts the enhancing effect of NF-κBp50 on c-maf binding activity. Whether this effect requires other cofactor(s) remains to be determined.

The finding that forced expression of NF-κBp50 alone induces transcription of the IL-4 gene seems to indicate that NF-κBp50 activates other target genes involved in Th2 polarization. In this respect, data obtained from NF-κBp50-deficient mice demonstrate that p50 plays a pivotal role in the induction of GATA-3 expression, another master Th2 transcription factor (32). Taken together, these results provide evidence of nonoverlapping functions for NF-κB family members in the development of Th2 immune response. It is interesting to note that independent works have reported an increase of IL-13 but not IL-4 in MCNS (9, 10). Yet, c-maf does not induce IL-13, and c-maf-deficient mice express normal levels of IL-13 (33). Although GATA-3 transactivates IL-4 and IL-13 genes, the role of GATA-3 in IL-13 increase during MCNS relapses is unlikely, because the inhibition of NF-κB in Th2 cells did not affect GATA-3 expression but significantly reduces IL-13 production (32). An interpretation of this finding is that NF-κB is directly implicated in the transcription of the IL-13 gene. In support of this view, two NF-κB-responsive elements have been localized on the IL-13 promoter (32).

Recent data strongly suggest that c-maf increases the IL-10 expression (33). This is not the case in our patients in whom we did not detect a significant difference in IL-10 mRNA levels between relapse and remission. Importantly, normal subjects exhibited similar IL-10 levels even though their expression of c-maf-S was very modest.

The strong induction of c-maf-S in MCNS, together with its functional characteristics, regardless of decoupling of IL-4 production, suggests that the target genes of c-maf in MCNS are still to be discovered. Identification of gene(s) activated by c-maf in MCNS likely would be an important step toward defining the pathogenesis of this disease.

We are grateful to Vincenzo Casolaro for providing the IL-4-CAT promoter constructs and Qin Chen for c-maf expression vector. We thank Y. Laperche and J. P. Farcet for critical review of the manuscript. We are indebted to Drs. P. M. Ronco, V. Baudouin, P. Niaudet, M. Broyer, F. Bouissou, B. Boudailliez, and our numerous colleagues, for providing us with blood samples and clinical information, as well as for their support and advice.

1

This work was supported in part by Association pour l’Utilisation du Rein Artificiel (Paris, France), Novartis, Université Paris XII, and an APEX grant from Institut National de la Santé et de la Recherche Médicale. D.S. is a recipient of Programme AVENIR (Institut National de la Santé et de la Recherche Médicale).

3

Abbreviations used in this paper: MCNS, minimal change nephrotic syndrome; Mare, Maf recognition element; MN, membranous nephropathy; c-maf-L, c-maf-long; c-maf-S, c-maf-short; DAPI, 4′,6′-diamidino-2-phenylindole; wt, wild type; mt, mutant.

1
Niaudet, P..
1993
. Nephrotic syndrome in children.
Curr. Opin. Pediatr.
5
:
174
.
2
Michael, A. F., R. H. McLean, L. P. Roy, N. G. Westberg, J. R. Hoyer, A. J. Fish, R. L. Vernier.
1973
. Immunologic aspects of the nephrotic syndrome.
Kidney Int.
3
:
105
.
3
Glassock, R. J., S. G. Adler, H. J. Ward, A. H. Cohen.
1991
. Minimal change disease in primary glomerular diseases. B. M. Brenner, and F. C. Rector, eds. 4th Ed.In
The Kidney
Vol. 1
:
1224
. Saunders, Philadelphia.
4
Schnaper, H. W..
1989
. The immune system in minimal change nephrotic syndrome.
Pediatr. Nephrol.
3
:
101
.
5
Boulton Jones, J. M., I. Tulloch, B. Dore, A. McLay.
1983
. Changes in the glomerular capillary wall induced by lymphocyte products and serum of nephrotic patients.
Clin. Nephrol.
20
:
72
.
6
Yoshizawa, N., Y. Kusumi, K. Matsumoto, S. Oshima, A. Takeuchi, O. Kawamura, T. Kubota, S. Kondo, H. Niwa.
1989
. Studies of a glomerular permeability factor in patients with minimal-change nephrotic syndrome.
Nephron.
51
:
370
.
7
Tanaka, R., N. Yoshikawa, H. Nakamura, H. Ito.
1992
. Infusion of peripheral blood mononuclear cell products from nephrotic children increases albuminuria in rats.
Nephron.
60
:
35
.
8
Sahali, D., A. Pawlak, A. Valanciute, P. Grimbert, P. Lang, P. Remy, A. Bensman, G. Guellen.
2002
. A novel approach to investigation of the pathogenesis of active minimal-change nephrotic syndrome using subtracted cDNA library screening.
J. Am. Soc. Nephrol.
13
:
1238
.
9
Kimata, H., M. Fujimoto, K. Furusho.
1995
. Involvement of interleukin (IL)-13, but not IL-4, in spontaneous IgE and IgG4 production in nephrotic syndrome.
Eur. J. Immunol.
25
:
1497
.
10
Yap, H. K., W. Cheung, B. Murugasu, S. K. Sim, C. C. Seah, S. C. Jordan.
1999
. Th1 and Th2 cytokine mRNA profiles in childhood nephrotic syndrome: evidence for increased IL-13 mRNA expression in relapse.
J. Am. Soc. Nephrol.
10
:
529
.
11
Fodor, P., M. T. Saitua, E. Rodriguez, B. Gonzalez, L. Schlesinger.
1982
. T-cell dysfunction in minimal-change nephrotic syndrome of childhood.
Am. J. Dis. Child.
136
:
713
.
12
Blank, V., N. C. Andrews.
1997
. The Maf transcription factors: regulators of differentiation.
Trends Biochem. Sci.
22
:
437
.
13
Chesi, M., P. L. Bergsagel, O. O. Shonukan, M. L. Martelli, L. A. Brents, T. Chen, E. Schrock, T. Ried, W. M. Kuehl.
1998
. Frequent dysregulation of the c-maf proto-oncogene at 16q23 by translocation to an Ig locus in multiple myeloma.
Blood
91
:
4457
.
14
Kataoka, K., M. Noda, M. Nishizawa.
1994
. Maf nuclear oncoprotein recognizes sequences related to an AP-1 site and forms heterodimers with both Fos and Jun.
Mol. Cell. Biol.
14
:
700
.
15
Ho, I. C., M. R. Hodge, J. W. Rooney, L. H. Glimcher.
1996
. The proto-oncogene c-maf is responsible for tissue-specific expression of interleukin-4.
Cell
85
:
973
.
16
Ho, I. C., L. H. Glimcher.
2002
. Transcription: tantalizing times for T cells.
Cell
109
:
S109
.
17
Kim, J. I., T. Li, I. C. Ho, M. J. Grusby, L. H. Glimcher.
1999
. Requirement for the c-Maf transcription factor in crystallin gene regulation and lens development.
Proc. Natl. Acad. Sci. USA
96
:
3781
.
18
Ranganath, S., K. M. Murphy.
2001
. Structure and specificity of GATA proteins in Th2 development.
Mol. Cell. Biol.
21
:
2716
.
19
Sahali, D., A. Pawlak, S. Le Gouvello, P. Lang, A. Valanciute, P. Remy, C. Loirat, P. Niaudet, A. Bensman, G. Guellaen.
2001
. Transcriptional and post-transcriptional alterations of IκBα in active minimal-change nephrotic syndrome.
J. Am. Soc. Nephrol.
12
:
1648
.
20
Anonymous, .
1981
. The primary nephrotic syndrome in children: identification of patients with minimal change nephrotic syndrome from initial response to prednisone: a report of the International Study of Kidney Disease in Children.
J. Pediatr.
98
:
561
.
21
Thanos, D., T. Maniatis.
1992
. The high mobility group protein HMG I(Y) is required for NF-κB-dependent virus induction of the human IFN-β gene.
Cell
71
:
777
.
22
Chen, Q., D. H. Dowhan, D. Liang, D. D. Moore, P. A. Overbeek.
2002
. CREB-binding protein/p300 co-activation of crystallin gene expression.
J. Biol. Chem.
277
:
24081
.
23
Philipsen, S., G. Suske.
1999
. A tale of three fingers: the family of mammalian Sp/XKLF transcription factors.
Nucleic Acids Res.
27
:
2991
.
24
Hedge, S. P., A. Kumar, C. Kurschner, L. H. Shapiro.
1998
. c-Maf interacts with c-Myb to regulate transcription of an early myeloid gene during differentiation.
Mol. Cell. Biol.
18
:
2729
.
25
Daniel, V., Y. Trautmann, M. Konrad, A. Nayir, K. Scharer.
1997
. T-lymphocyte populations, cytokines and other growth factors in serum and urine of children with idiopathic nephrotic syndrome.
Clin. Nephrol.
47
:
289
.
26
Koyama, A., M. Fujisaki, M. Kobayashi, M. Igarashi, M. Narita.
1991
. A glomerular permeability factor produced by human T cell hybridomas.
Kidney Int.
40
:
453
.
27
Hulton, S. A., T. J. Neuhaus, R. E. Callard, M. J. Dillon, T. M. Barratt.
1997
. Circulating interleukin 2 receptor (IL-2R) in nephrotic syndrome.
Kidney Int. Suppl.
58
:
S83
.
28
Grimbert, P., A. Valanciute, V. Audard, A. Pawlak, S. Le Gouvelo, P. Lang, P. Niaudet, A. Bensman, G. Guellaen, D. Sahali.
2003
. Truncation of C-mip (Tc-mip), a new proximal signaling protein, induces c-maf Th2 transcription factor and cytoskeleton reorganization.
J. Exp. Med.
198
:
797
.
29
Cao, S., J. Liu, M. Chesi, P. L. Bergsagel, I. C. Ho, R. P. Donnelly, X. Ma.
2002
. Differential regulation of IL-12 and IL-10 gene expression in macrophages by the basic leucine zipper transcription factor c-Maf fibrosarcoma.
J. Immunol.
169
:
5715
.
30
Ho, I. C., D. Lo, L. H. Glimcher.
1998
. c-maf promotes T helper cell type 2 (Th2) and attenuates Th1 differentiation by both interleukin 4-dependent and -independent mechanisms.
J. Exp. Med.
188
:
1859
.
31
Almawi, W. Y., O. K. Melemedjian.
2002
. Negative regulation of nuclear factor-κB activation and function by glucocorticoids.
J. Mol. Endocrinol.
28
:
69
.
32
Das, J., C. H. Chen, L. Yang, L. Cohn, P. Ray, A. Ray.
2001
. A critical role for NF-κB in GATA3 expression and TH2 differentiation in allergic airway inflammation.
Nat. Immunol.
2
:
45
.
33
Kishikawa, H., J. Sun, A. Choi, S. C. Miaw, I. C. Ho.
2001
. The cell type-specific expression of the murine IL-13 gene is regulated by GATA-3.
J. Immunol.
167
:
4414
.